Research On Social-Relation Based Routing Algorithms In DTN

The DTN (Delay/Disruption Tolerant Networks) is an abstractnetwork model which is different from traditional networks. It transmitsmessages based on the communication opportunities by node mobility,and utilizes the unique "store-carry-and-forward" transmissioncharacteristic. In such networks, continuous end-to-end connectivity maynot be possible. Messages are opportunistically routed when mobile nodesencounter each other. Therefore, DTN is also a class of opportunisticnetworks, whose totally different networking attracts extensive attentionsfrom both researchers and developers.The topology of DTN is not stable, and the communication statebetween two nodes can be changed timely under nodes' mobility, whichmake the transmission latency and delivery of messages uncontrolleable.As a result, traditional network models and algorithms are not suitable forthe challenging environment in DTN, such as vehicular DTNs,pocket-switched networks, scientific expedition networks, rural networks,warfield networks, etc. In this context, studies on DTN architectures androuting algorithms have been extensively conducted. Nonetheless,systematic research on DTN routing algorithm from the perspective ofsocial relations among nodes is still open and far from a perfecttheoretical system, though with some successful results.This paper constructs DTN models and designs routing algorithms mainly from the perspective of social relations. Through detailed analysisof the characteristics of MaxProp routing algorithm which has beensuccessfully applied to the realistic DTN network environment, weanalyze its advantages and disadvantages, and then innovatively come upwith cost based PriCost and Markov Chain based MCR routingalgorithms. By simulation to measure comprehensive indicators (i.e.delivery ratio, average latency, average hopcount and overhead ratio, etc.),experimental results demonstrate both algorithms outperform traditionalalgorithms like Maxprop and Prophet.